Meeting Banner
Abstract #1795

Radiomics Driven Image Analysis and Coregistration Scheme to Identify DCE MRI Markers for Microvascular Density

Asha Singanamalli1, Rachel Sparks1, Mirabela Rusu1, Natalie Shih2, Amy Ziober2, John Tomaszewski3, Mark Rosen4, Michael Feldman2, Anant Madabhushi1

1Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States; 2Department of Pathology, University of Pennsylvania, Philadelphia, PA, United States; 3Department of Pathology & Anatomical Sciences, University of Buffalo, University of Buffalo, Buffalo, United States; 4Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Prostate cancer is the second leading cause of cancer mortality in men. Measures such as gleason scores and microvascular density (MVD) are well established indicators of CaP outcome. However, invasive procedures are necessary to obtain these measures. In this study, we seek to identify DCE MRI perfusion parameters that correlate with MVD using sophisticated image analysis and registration methods in order to establish non-invasive methods of assessing tumor outcome.

Keywords

accounted accounts allowed anatomical annotated assessing assessment assuming biomedical biopsy blue buffalo cancer cause chose coefficient color combines confirmed contiguous correlate correlation correspondences curve curves cuts defined density determined driven emerged energy engineering enhanced enhancement establish established expert explained fact feature features fraction frequency function functional general graph highest histological histology house hypothesize identify indicate indicators intensity interactive invasive john kinetic landmarks latter learning linearity localize location machine mark marker markers measures medical minimized modalities modality mortality mount moving necessary nonlinear nonlinearity normalized outcome overlap particular pathologist pathology patients pattern peak permeability permeation plate plot preliminary prior procedures program promising prostate quadrants qualitatively radical radiologist radiology rank recognition registered registration relationship representative reserve sciences score scores sectioned sections segment segmented selects serve significantly since slice slices sophisticated space sparks specimens splines stain stained stitched suggest suggests swatch table thin transactions trends tumor tumors underwent uptake user vascular washout western white whole